Largest exome-sequencing meta-analysis to date (28,898 cases, 103,041 controls, 3,444 trios) identifies STAG1 and ZNF136 at exome-wide significance and six additional genes at FDR<5%, highlighting roles for chromatin organisation and GABAergic signalling.
0:15Late night screams. Welcome to Base by Base, the papercast that brings genomics to you, wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Glad to be here for this one.
0:29So to start off today, I want you to imagine a massive symphony orchestra. Okay, a symphony. Right. So we are talking 100s of musicians, and they're all playing together in what should be a perfectly coordinated performance.
0:44But as you sit there and listen, you notice this really profound dissonance. Like something is just deeply wrong with the music. Exactly. The whole acoustic structure of the performance is fundamentally altered.
0:59Now, you wouldn't diagnose the problem by just looking at the conductor. No, of course not. have to get in there and isolate the specific chairs. Yeah, you have to find the individual broken instruments to see exactly which structural flaws are throwing the entire peas into disarray.
1:12And, you know, that challenge amplifies immensely when that symphony is the human brain. And the resulting dissonance manifests as schizophrenia. I mean, it is a highly complex syndrome, and it's defined by these profound disruptions to cognition, perception, and behavior.
1:29Which is why we're doing this deep dive today. Because for decades, we've relied on antipsychotic medications. But well, they often fail to adequately resolve the core cognitive symptoms. Yeah, because they are essentially blunt instruments, they target these broad receptor networks rather than the root etiology, the actual broken strings in the orchestra, so to speak.
1:50Right, which makes isolating the exact genetic typos driving the disorder so critical. Because if we can map the specific molecular instruments causing the dissonance, we fundamentally shift how we approach the biology of the condition.
2:04We move from just, you know, symptom management to targeted mechanism-based interventions. And that underlying goal brings us to some truly sophisticated work in psychiatric genetics that we are dissecting today.
2:15Yes, it does. So today we celebrate the work of researchers at Cardiff University, including Sophie Elchick, Elliot Reese, and their collaborators, who have advanced our understanding of the genetic architecture of schizophrenia.
2:28It's an incredible body of work. It really is. This was published in 2025 in nature communications. But to really appreciate the magnitude of what this team found. We 1st have to explain the puzzle they were trying to solve.
2:41Right. The statistical wall the field had hit. So schizophrenia is highly polygenic. That means 1000s of genetic variations contribute to it. Now, common genetic variants explain about 24% of the risk.
2:54Okay, 24%. But wait, hold on. If common variants explain 24% of the risk, why are scientists hunting so hard for these rare variants? I think they only explain about 5% of the liability right? Yeah, ultra rare coding variants and copy number variants explain about 5%.
3:11So why invest so much time and money, hunting down a measly 5% when you have 24% sitting right there? It's a great question. It seems counterintuitive until you look at the biological clarity those variants provide.
3:25What do you mean by clarity? Well, the common variants are diffuse. They create a sort of background risk profile, mostly sitting in non-coding regions, so they're notoriously hard to map to specific causative genes, but the rare variants, they act differently.
3:40How so? They are concentrated in a network of roughly 3000 highly constrained genes. Meaning they are incredibly intolerant to mutations. Evolution actively protects these genes. So if a mutation does happen here, it has a massive direct effect.
3:58It points directly to specific causal genes and disease mechanisms in a way that common variants just do not. Oh, wow. Okay, so it's like finding a smoking gun instead of just circumstantial evidence. Exactly.
4:10Previous landmark research, like the schema study, had already started this, but they only found 12 genes at an XM wide level of certainty. And 12 genes just isn't enough to understand the whole network.
4:21Right. We needed more. But the problem wasn't the science. It was the statistical power. You simply cannot find these ultra rare typos with a small group of people. So how do you find them? I mean, you'd need a monstrously large library of genome.
4:35You really do. And that is where the Cardiff team came in. They generated whole X-one sequencing data for a completely new sample. How big was the new sample? We're talking 4,650 cases and 5,719 controls.
4:49Just for this one study. Yep. And then they meta-analyze this with the published data from that earlier schema study. Okay, so they combined everything. What does the final math look like on that? It is staggering.
5:00It makes this the largest whole X-Sum sequencing meta-analysis of schizophrenia to date. The total comes out to 28,000,898 cases. Wow. 103,041 controls, and 3444 pro band parent trios. That is almost 130,000 genomes.
5:18For you listening, imagine trying to proofread 1000000s of books just to find a specific missing letter. That's essentially what they did. Yeah, and they were specifically hunting for 2 types of typos.
5:26First, protein truncating variants or PTVs. And those are the mutations that prematurely stop protein production, right? Like a sentence that just ends. Exactly. Abruptly stops. The protein is just severely truncated and nonfunctional.
5:41The 2nd type they looked for were damaging the sense variants. Misense variant. So that's where a single wrong amino acid is swapped in. The protein gets built, but it has a bad part. Right. And to figure out how bad that part is, they scored it using an MPC metric.
5:57It basically measures the badness of the mis sense variant, prioritizing mutations that disrupt really important, highly conserved regions. Okay, so this massive proofreading effort with almost 130,000 genomes.
6:11Did it pay off? Let's get to the newly discovered genetic culprits. It absolutely paid off. Two novel wrist genes hit the absolute gold standard of XM wide significance. And those are... 7G1 and ZNF 136.
6:25Nice. Now, weren't those on the radar before? They were. They were highly suspected in the earlier data, but this new massive data set finally pushed them over the statistical finish line. That's incredible.
6:35And I know there was a supporting cast of genes too, right? Yes. The increased sample size identified 6 additional genes at a false discovery rate of less than 5%. Let me guess, a bunch of random letters and numbers.
6:46You know it. We've got SLC 6A1, PCLO, ZMYND 11, BSCL2, KLC1, and CGREF1. Okay, I want to point out something fascinating from the data regarding 2 of those. For SLC 6A1 and KLC1, the association was driven entirely by those misensurians we talked about.
7:05Not by the protein completely truncating. It's not about the protein disappearing. It's about it being built slightly wrong. Exactly. It might preserve the overall protein, but disrupt a vital interaction domain.
7:18It really highlights how delicate these cellular pathways are. So knowing the names of the genes is just the start. What do these genes actually build or break in the brain? Let's unpack the biology, starting with stage one?
7:29So stig one encodes a subunit of cohesion. Cohesion. Right, which is required for 3D genome organization. It basically extrudes loops of DNA to form topologically associating domains or tads. Okay, so think of it like molecular origami.
7:43If the DNA isn't folded correctly, if the cohesion complex is compromised by a typo in stag one, the final 3D shape is ruined, meaning genes are turned on or off at the wrong times because they're physically in the wrong spot, this directly points to disrupted chromatin organization as a root cause of schizophrenia.
8:02It's profound, isn't it? We aren't just talking about a chemical imbalance. We're talking about the physical 3D architecture of the genome being altered. What about ZNF 136. That was the other big one.
8:12Yeah, so ZenF 136 encodes a transcriptional repressor. But here is the weird part. Unlike almost every other schizophrenia risk gene, it shows no evolutionary constraint against protein truncating variants.
8:26Wait, really? No constraint, but you said earlier that genes causing severe disorders are usually highly constrained because evolution weeds out the mutations. Exactly. It is a huge scientific mystery.
8:36Why is this repressor gene an exception? Maybe its function is buffered by other proteins, or maybe the severe effects only show up under certain environmental stressors we just don't know yet? That is wild.
8:46Okay, let's look at SLC 6A1. The one driven by those misense variants. So this one encodes a GABBA transporter, specifically JTor one. And GABA is a neurotransmitter. Right. The main inhibitory one in the brain.
8:59The misins variants in this gene cause haplo insufficiency, which basically means reduced GABA uptake at the inhibitory synapses. So the chemical signaling is perturbed. The brakes in the brain aren't working, right?
9:11Exactly. It anchors that old theory about perturbed gabergic signaling to hard, exomic data. But, and this is important. Several of these genes are also linked to other disorders. Oh, like pleatropy? Yeah.
9:22For example, SLC6A1 is also strongly implicated in autism spectrum disorder, developmental disorders and epilepsy. And stag one is tied to developmental disorders too. Wait, hold on. If the exact same gene causes epilepsy, autism, and schizophrenia, does that mean these diseases are the same thing?
9:41Are we just giving different names to the same biological problem? No, at all. And that's where we have to be really precise. It's about different variants within that exact same gene, having entirely distinct functional effects.
9:52Oh, so it depends on where the typo is. Exactly. A severe truncating mutation that completely kills the transporter might cause early onset epilepsy, but a milder misense typo that only slightly reduces efficiency.
10:06That might subtly change brain networks and contribute to schizophrenia much later in life. Ah, okay. That makes a lot of sense. The specific nature of the perturbation dictates the trajectory of the disorder.
10:18Precisely. And there's another really exciting piece to this. Stag one and KLC one physically overlap with genomic regions that were previously identified in common variant studies. The GWS studies. Yes. Why is that important?
10:31Because that overlap is the Holy Grail. It bridges the rare, high impact mutations with the common, low impact signals in the broader population, it makes these genes perfect targets for developing future cellular and animal models.
10:45Because if you study the rare, severe mutation and a lab model, you are essentially studying a magnified version of the exact same pathway that affects the general population. You hit the nail on the head.
10:55That is incredible. Yeah. But I imagine there are some limitations to a study like this, right? Even with 130,000 genomes. Unfortunately, yes. The biggest issue is that the study lacked deep longitudinal clinical data.
11:09Meaning they had the genetics, but not the detailed patient histories. Right. We have the genetic architecture, but we cannot yet link these specific genes to specific patient symptoms. We don't know if a stag one mutation correlates with negative symptoms or cognitive deficits or whatever else.
11:26Which is what you'd need for true precision medicine. Exactly. And the other major limitation is diversity. The vast majority of these genetic samples were from individuals of European ancestry. Oh, yeah, that's a persistent problem in genomic studies.
11:39It really is. The study highlights a desperate need for more diverse non-European genetic samples to ensure equitable progress in precision psychiatry. You can't fully map the global landscape of a disorder if you are only looking at one demographic.
11:53So true. We need the whole picture. Well, stepping back from the molecular origami and the GABA transporters, let's look at the big picture for you listening today. Yeah, let's synthesize this. The largest XM sequencing study to date, successfully utilized nearly 130,000 genomes to definitively pin down new risk genes for schizophrenia.
12:14By identifying genes like stag one and SLC6A1, scientists now have concrete evidence that the physical 3D folding of DNA and the chemical signaling in our brains are critical pathways in the development of the disorder.
12:27What does this mean for the future of psychiatric medicine if we can start treating the specific genetic folding errors instead of just the psychological symptoms? This episode was based on an open access article under the CCBY 4.0 license.
12:40You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating. If you'd like to support our work, use the donation link in the description.
12:52Now stay with us for an original track created, especially for this episode, and inspired by the article you've just heard about. Thanks for listening, and join us next time as we explore more science base by base.
13:19Late night screams We come the letters down to bone. Tiny brakes where shadows grow. In lines of life we call our own. Not just one spark, not one name. Its patterns hiding in the frame. Pull the thread, feel it pull back, stitched up sky with subtle cracks.
13:49Hear the boss in the silent Group made in nuts in the garbage and rare little switches, heavy like stone. Fault lines, code was shown, but we're learning where the dark begins. So we can build, build, build, build the light back in, in, in.
14:33Clipped off protein, missing, and bent up message, miss, and misread. Some signals cross where they shouldn't land. Some brakes don't bite inside the head. And come and waves meet rare extremes. On overlapping scenes of genes.
14:55Not destiny, not blame, not curse. Just much to help us do the work. Follow up fire models to grow. From what the egg zones. Dare to shine. Here, post in the silent room. Coherent, holding and coming loose.
15:26Transport is firing out of dome. Fall lines and code where shown all the names rise up like neon signs, pointing the pathways, we can redesign from stitched up nights, do a clear view. We find the threat.
15:47And me pull it through
